The natural history of clinically localized prostate cancer is controversial. However, the data that is available suggests outcomes are good for men with well and moderately differentiated disease, but relatively poor for men with poorly differentiated cancer. For example, in one synthesis of individual-patient level data from six series, ten-year prostate cancer-specific mortality for men with well, moderate, and poorly differentiated disease was estimated at 13%, 13%, and 66%, respectively. Tumor grade is the dominant predictor of prognosis. Few studies report on risks of morbidity due to local cancer progression; the data available suggest this risk is relatively low. Cancer-specific mortality figures overestimate the impact of disease among older men with competing causes of mortality. Given the available data on prognosis of men with localized prostate cancer not treated for cure, clinical trials designed to measure the effectiveness of aggressive therapy will need to be large.